Allocating scarce resources is tough. Someone should study that...
And while I haven't studied the hotel room situation, here are some off the top of the head thoughts.
One way to allocate resources of course is by price: we could auction off the nicest and most convenient rooms to the highest bidders. We economists all know the good properties that mechanism has, and most of us, if we stop to think about it, know about some of the bad properties as well. For better or worse, the AEA has concentrated on getting uniformly low room rates (which is not so hard to do when you are filling hotels the first weekend in January...)
So, how about a matching mechanism, in which prices don't do the heavy lifting of helping the market clear, but we ask people about their preferences for different kinds of rooms?
The potential efficiency gains from that kind of endeavor arise when people have heterogeneous preferences. If most people have quite similar preferences (1st choice=main conference hotel, lesser choices ranked by distance from main conference hotel... with some variation for suites versus singles, etc.), then many incentive compatible matching mechanisms end up being equivalent to some sort of serial dictatorship. (Of course, the degree of heterogeneity of (cardinal) preferences is an empirical question, for which I have no data.)
So now we're talking about who should get to choose first. And the system we have lets that be endogenous--the most desirable worms go to the early birds.
There could be better ways, but there aren't wonderful ways to allocate commonly-valued resources when you're reluctant to use prices...